Cell lines and therapeutic antibodies

Murine cancer cell lines 4T1 (breast cancer), EMT-6 (breast cancer), H22 (hepatocellular carcinoma) were cultured with RPMI-1640 (Gibco) containing 10% fetal bovine serum (FBS) (Excell). Human cancer cell line A549 (lung cancer) was cultured with F12-K (21127022, Gibco) containing 10% FBS. Human cancer cell line MDA-MB-231 (breast cancer) was cultured with DMEM (Gibco) containing 10% FBS. Murine T cell lines CTLL-2 and HT-2 were cultured in RPMI-1640 (ATCC modification, containing glutathione and vitamins) (A10491-01, Gibco) with 10% FBS and 200 IU/ml interleukin-2 (IL-2, Beijing Fourrings). 293-NFAT was cultured with MEM (12561-056, Gibco) containing 10% FBS. HUVEC (human umbilical vein endothelial cell) was cultured with ECM (1001, Sciencell) containing 10% FBS.

The therapeutic antibodies including anti-VEGF, anti-TGF-β, anti-PD-1, Y332D (anti-VEGF/TGF-β bispecific antibody) and isotype control antibody (human IgG) in this study were provided by Wuhan YZY Biopharma.

Sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE)

The prepared Y332D was analyzed by SDS-PAGE with Coomassie Brilliant Blue staining. To verify the molecular weight of Y332D, non-reduced and reduced SDS-PAGE were performed. The non-reduced sample was prepared by mixing the protein with 5 μl of 0.5 M 2-Iodoacetamide (IAM) solution and incubating for 30 min away from light, then mixing the sample with 10 μl of NR Sample Buffer and incubating for 5 min at 60 °C. The reduced sample was prepared by mixing the protein with 10 μl of R Sample Buffer and incubating for 10 min at 70 °C. The sample was separated via 4–20% SurePAGE. After the SDS-PAGE gel was stained with Coomassie Brilliant Blue and decolorization, the image was captured with ChemiDoc MP Imaging system (Bio-Rad).

Capillary electrophoresis with sodium dodecylsulfate (CE-SDS)

To verify the purity of Y332D, non-reduced and reduced CE-SDS were conducted. For non-reduced CD-SDS, 100 μg Y332D was mixed with 1 μl 10 kD Internal Standard and 5 μl 0.5 mol/l IAM solution, incubated at room temperature and protected from light for 30 min, and then the SDS-MW Sample buffer was added to 101 μl. The mixture was heated at 60 °C for 5 min and cooled at room temperature for 3 min. For reduced CE-SDS, 100 μg Y332D was mixed with SDS-MW Sample Buffer to make a total volume of 95 μl, and then 1 μl 10 kD Internal Standard and 5 μl β-mercaptoethanol were added. The prepared mixture was heated at 70 °C for 10 min and cooled at room temperature for 3 min. UV of migratory proteins was monitored at 214 nm using Beckman PA800 Plus.

Surface plasmon resonance (SPR)

Protein A chip was used to immobilize Y332D by capture method, and VEGFA and TGF-β1 antigens were used as analytes to detect the kinetics and affinity of their binding to Y332D. Y332D was diluted to 2 μg/ml, and VEGFA and TGF-β1 antigens were diluted to 10 nM and 100 nM, respectively. Then, the “Biacore T200 Control Software” was run to determine the assay conditions as capture concentration of 2 μg/ml for Y332D, flow rate of 10 μl/min, and binding for 120 s. The VEGFA and TGF-β1 antigens binding parameters were 30 μl/min flow rate, 120 s binding and 600 s dissociation. The starting concentration of VEGFA antigen and Y332D were 5 nM, based on which a twofold serially dilution was applied. The starting concentration of TGF-β1 antigen and Y332D were 10 nM, based on which a twofold serially dilution was applied. The regeneration conditions were flow rate of 10 μl/min and binding for 60 s, and the regeneration reagent was 10 mM Glycine–HCl, pH 1.5. Sample detection conditions were set and 2–1 channels were selected for sample detection. After the assay was completed, the data were fitted using the “1:1 Binding” in Biacore T200 Evaluation Software, and the dissociation equilibrium constant (Kd) was calculated.

Enzyme-linked immunosorbent assay (ELISA)

TGF-β1 (200 ng per well), VEGFA (100 ng per well) were coated in 96-well flat-bottom plates (3599, Costar) overnight at 4 °C. The next day, the plates were washed three times with PBS containing 0.05% Tween 20 and then blocked with 5% Bovine Serum Albumin (BSA, BSAS1.0, Bovogen) for 3 h. Then, serially diluted Y332D or controls (100 μl per well) were added to the plates and incubated at 37 °C for 2 h. After the plates were washed, anti-hIgG-HRP (1:5000, A80-319P, Bethyl) was added and incubated at 37 °C for 1 h. Subsequently, TMB chromogenic solution (P0209, Beyotime) was added to the washed plates, and the reaction was terminated with TMB chromogenic termination solution (P0215, Beyotime). Finally, the absorbance values were read at 450 nm.

Preparation of VEGFA-Biotin with Biotin Labeling Kit-NH2 (LK03, DoJindo): 100 µl WS buffer and 100 µg VEGFA protein solution were mixed and centrifuged at 8000 g for 10 min. The prepared mixture was added with 8 µl NH2-reactive biotin and 100 µl Reaction buffer and incubated at 37 °C for 30 min. Subsequently, the labeled VEGFA protein was washed three times with WS buffer and set aside with 100 µl WS buffer.

To detect the simultaneous binding affinity of Y332D, we performed double-antigen sandwich ELISA. TGF-β1 (200 ng per well) was coated in 96-well flat-bottom plates overnight at 4 °C. The next day, the plates were washed three times with PBS containing 0.05% Tween 20 and then blocked with 5% BSA for 3 h. Subsequently, serially diluted Y322D or controls (100 μl per well) were added to the plates and incubated at 37 °C for 2 h. Then, the plates were washed, and 100 µl prepared VEGFA-Biotin (20 ng per well) and peroxidase-conjugated streptavidin (1:5000, SA00001-0, Proteintech) were added. Finally, the plates were added with TMB chromogenic substrate and the reaction was terminated followed by the detection of absorbance values at 450 nm.

CCK-8 assay

TGF-β1 could impede the proliferation of IL-2-dependent murine T cells [41]. CCK-8 assay was performed to explore the antagonistic effect of Y332D on the activity of TGF-β. 1 × 103 CTLL-2 and HT-2 cells were seeded in 96-well plates. Then, 5 ng/ml TGF-β1 and 106 pM antibodies were added. Cell viability was continuously monitored by CCK-8 reagent (10 μl per well, Dojindo) within one week after treatment.

In vitro cytokine detection

To investigate the effect of Y332D on the alteration of TGF-β1-caused cytokine secretion during T cell activation, we conducted multi-cytokine assay using Cytometric Bead Array (CBA) Mouse Th1/Th2/Th17 Cytokine Kit (560485, BD Biosciences). Murine T cells were obtained from the isolation of splenocytes from C57BL/6 mice by Dynabeads™ Untouched™ Mouse T Cells Kit (11413D, Invitrogen). T cells (1 × 106/ml) supplemented with anti-CD28 (3 μg/ml, 102116, Biolegend), TGF-β1 (5 ng/ml) and 106 pM antibodies were cultured in 96 well flat-bottom plates precoated with anti-CD3 (3 μg/ml, 100302, Biolegend). After 4 days, the cellular supernatants were harvested and CBA Mouse Th1/Th2/Th17 Cytokine Kit was used to measure cytokines concentration.

SBE4 luciferase reporter assay

3 × 104 A549 or MDA-MB-231 cells were plated in 96 well flat-bottom plates and incubated at 37 °C overnight. The next day, cells were transiently transfected with 0.2 μg SBE4 luciferase reporter plasmid for each well by Lipofectamine 2000. Cells were starved 24 h post-transfection and treated with 10 ng/ml TGF-β1 and 106 pM Y332D or controls for 24 h. Luminescence was detected using Bio-Lite™ Luciferase Assay System (DD1201, Vazyme).

Western blotting

Tumor cells were lysed using RIPA buffer (P0013B, Beyotime). The supernatant was collected after centrifugation at 14,000 rpm for 15 min at 4 °C, and the total protein concentration was measured using BCA assay kit (P0010S, Beyotime). 30 µg protein sample was separated using SDS-PAGE gel (NP0321BOX, Life Tech) and transferred to a polyvinylidene fluoride membrane (ISEQ00010, Millipore). Then, the membranes were blocked with 5% BSA for 1 h and incubated with the following primary antibodies: anti-N-cadherin (1:1000, 13116, CST), anti-Vimentin (1:1000, 5741, CST), anti-GAPDH (1:1000, 5174, CST), and anti-β-Actin (1:5000, AF7018, Affinity) at 4 °C overnight. The next day, the membranes were incubated with the secondary antibodies Goat-anti-rabbit-IgG-HRP (1:2000, 7074, CST) for 1 h. SuperSignal™ West Pico PLUS (34577, Thermo Scientific) was used for visualization, and the G:BOX Chemi X system was used for signal detection.

Transwell migration and invasion assays

To measure the motility of tumor cells, transwell migration and invasion assays were performed using 24-well Transwell apparatus containing 6.5-mm polycarbonate membrane with 8-μm pore size inserts (3422, Corning) without or with Matrigel (354234, BD Biosciences). 4T1 and EMT-6 mammary tumor cells were cultured in RPIM-1640 with 1% FBS and treated with 5 ng/ml TGF-β1 plus 106 pM antibodies or untreated for 96 h. Then, about 5 × 104 cells in 100 µl RPIM-1640 supplemented with 1% FBS were seeded in the upper chambers. The lower chambers were added with 600 µl of RPIM-1640 containing 10% FBS. After incubation for 24 h, the migratory and invasive cells were fixed with 4% paraformaldehyde (P0099, Beyotime) and stained with 0.1% crystal violet (C0121, Beyotime). Cell migration and invasion were evaluated by counting the migrated or invasive tumor cells in 5 random fields.

NFAT luciferase reporter assay

NFAT is a transcription factor downstream of the VEGF/VEGFR2 pathway. HEK-293 cells overexpressing VEGFR2 were transfected with the lentiviral vectors carrying the NFAT and luciferase gene (NFAT-RE-Luci) to construct stable transfected cell lines 293-NFAT. 293-NFAT cells were seeded in the 96 well flat-bottom plates and cultured in 2% FBS-DMEM with VEGFA (20 ng/ml) and serially diluted antibodies for 6 h at 37 °C. After each well was added with 80 μl Bio-Glo™ Luciferase Assay System (PRG7941, Promega) for 15 min, the luminescence was detected.

Luminescent cell viability assay

The inhibitory effect of Y332D on VEGFA-promoted HUVEC proliferation was measured using luminescent cell viability assay. Briefly, 5 × 103 HUVEC were seeded in 96 well flat-bottom plates overnight at 37 °C. The next day, the medium in the plates was discarded and ECM premixing with VEGFA (50 ng/ml) and serially diluted antibodies or control for 30 min were added and incubated at 37 °C for 72 h. Subsequently, 100 μl Cell Counting-Lite 2.0 detection reagent (DD1101-01, Vazyme) was added to each well and incubated for 10 min at room temperature. Then, the chemiluminescence was detected.

HUVECs tube formation assay

HUVECs were pre-cultured in ECM containing 1% FBS for 24 h. The next day, 2 × 104 HUVECs were seeded in 96 well flat-bottom plates after plates were precoated with 50 μl Matrigel (354234, BD Biosciences) for 30 min at 37 °C. The cells were incubated in endothelial cell complete medium mixing with 100 ng/ml VEGFA and 106 pM antibodies or control for 12 h at 37 °C. Then, HUVECs were fixed with 4% paraformaldehyde for 15 min. The images of tube-like structures were captured with inverted microscope (Olympus).

Murine tumor models

The antitumor activities of Y332D and anti-PD-1 plus Y332D were explored in multiple murine tumor models, including H22, EMT-6, 4T1, AKT/Ras-driven murine hepatocellular carcinoma.

For H22, EMT-6 and AKT/Ras-driven murine hepatocellular carcinoma models, 8.7 mg/kg anti-PD-1 was administrated every two days by intraperitoneal injection for four times. Equivalent mole hIgG (8.7 mg/kg), anti-VEGF (8.7 mg/kg), anti-TGF-β (6 mg/kg), Y332D (10 mg/kg) were administrated on alternate days by intraperitoneal injection for six times. For 4T1 lung metastasis model, Equivalent mole hIgG (8.7 mg/kg), anti-VEGF (8.7 mg/kg), anti-TGF-β (6 mg/kg), Y332D (10 mg/kg) were administrated on alternate days by intraperitoneal injection for six times. Tumor volume (TV) of tumor-bearing mice was measured every other day or every two days. TV was calculated by the following formula: volume = length × width2 × 0.5. Mice were euthanatized when TV exceeded 2500 mm3 or the study ended.

Subcutaneous H22 model

5 × 105 H22 cells were inoculated subcutaneously in the right groin of BALB/c mice on day 0. On day 6 after inoculation, treatment was started when the TV of the tumor-bearing mice reached 50–100 mm3. All tumor-bearing mice were randomly divided into eight groups: Vehicle, anti-VEGF, anti-TGF-β, Y332D, anti-PD-1, anti-PD-1 plus anti-VEGF, anti-PD-1 plus anti-TGF-β, anti-PD-1 plus Y332D (6 mice for each group).

Lung metastatic 4T1 model

2 × 104 4T1 cells were inoculated in the right mammary fat pad of BALB/c mice on day 0. Mice were anesthetized and subcutaneous tumors were removed when the TV reached 200–300 mm3. Then, mice were randomly divided into four groups (Vehicle, anti-VEGF, anti-TGF-β, Y332D) according to tumor volume (8 mice for each group). On day 34 after inoculation, mice were euthanized and lung tissues were collected for H&E staining.

Orthotopic EMT‑6 model

5 × 105 EMT-6 cells were inoculated in the right mammary fat pad of BALB/c mice on day 0. On day 10 after inoculation, treatment was started when the TV of the tumor-bearing mice reached 100–150 mm3. All tumor-bearing mice were randomly divided into six groups: Vehicle, Y332D, anti-PD-1, anti-PD-1 plus anti-VEGF, anti-PD-1 plus anti-TGF-β, anti-PD-1 plus Y332D (6 mice for each group).

AKT/Ras-driven murine hepatocellular carcinoma model

Hydrodynamic injection was performed to establish AKT/Ras-driven murine hepatocellular carcinoma model [42]. In brief, 5 μg of the plasmid encoding myr-AKT1 and/or 25 μg of the plasmid encoding NRasV12 along with 2 μg sleeping beauty transposase were diluted in 2 ml saline (0.9% NaCl), filtered through 0.22 μm filter, and injected into the lateral tail vein of 6 to 8-week-old C57BL/6 mice in 5–7 s.

Immunofluorescent (IF) staining

Freshly isolated tumor tissues were fixed in 10% neutral formalin for 48 h. The fixed tissues were then dehydrated, paraffin-embedded, sectioned and transferred to slides. IF staining based on tyramine signal amplification was performed according to the manufacturer's recommendations. The Multiplex Fluorescence Immunohistochemistry Kit-Four Color TSA-Rab-275 (10079100100, Panovue) was used in this assay. In addition, antibodies targeting E-cadherin (3195, CST), Vimentin (5741, CST), N-cadherin (13116, CST), α-SMA (19245, CST), CD31 (ab28364, Abcam), CD8 (98941, CST) were used in the assay. Images of IF were captured by fluorescence microscopy, previewed via Caseviewer software, and regions of interest were defined by two experienced pathologists. The quantitative analysis of IF images was conducted with ImageJ software.

Flow cytometry analysis of tumor-infiltrating lymphocytes (TILs)

Chopped tumor tissues were enzymatically digested using the dissociation buffer supplemented with 1 mg/ml Collagenase B (11088807001, Roche) and 1 mg/ml Hyaluronidase (H3506, Sigma-Aldrich) for 1 h at 37 °C. The prepared single-cell suspensions were filtered through 40-μm nylon meshes (352340, Corning) and centrifuged at 400 g for 5 min. Then, the centrifuged cells were treated with red blood cell lysis buffer (C3702, Beyotime). Subsequently, the cells were dyed with Fixable Viability Stain 780 (565388, BD), and Fc receptors were blocked with Ultra-LEAF™ Purified anti-mouse CD16/32 (101320, BioLegend). Cells were then fluorescently stained with the following detection antibodies: α-CD45 (103132, BioLegend), α-CD3 (100206, BioLegend), α-CD8 (100706, BioLegend), α-Ki67 (151215, BioLegend), α-CD69 (104536, BioLegend), α-CD25 (102012, BioLegend), α-CD107a (121629, BioLegend), α-Granzyme-B (372214, BioLegend), α-IFN-γ (505838, BioLegend). Brilliant Stain Buffer (563794, BD Biosciences) and True-Nuclear™ Transcription Factor Buffer Set (424401, BioLegend) were used in this assay. Flow cytometry was performed using Beckman CytoFLEX LX, and the data were analyzed by FlowJo_V10.

RNA-seq assay

At the end of H22 tumor treatment, four samples from each group were randomly selected for RNA-seq assay. The reference genome version was Mus Musculus (GRCm38/mm10). Total RNA was extracted by Trizol (Takara Bio) for cDNA library construction. Further deep sequencing was performed by Novogene (Beijing, China) via the Illumina Hiseq platform. Differentially expressed genes (DEGs) analysis was performed by R software with edgeR package and visualized by the heatmap package. DEG was identified as the gene with fold change over 2 and adjusted p-value less than 0.05. Immune signatures were designed based on the public lists [43]. Signature scores were defined by scaling the expression of all relevant genes within the signatures and calculating the mean value.

Quantitative real-time polymerase chain reaction (Q-RT-PCR)

Quantitative RT-PCR was performed to measure the expression patterns of immune-related genes in the transcriptome of tumor tissues. Gene-specific primers were designed and listed in Table 1. Total RNA was extracted and reverse-transcribed. Quantitative RT-PCR was run using the 7500 Real-Time PCR System (Applied Biosystems, USA) with iQ™ SYBR® Green Supermix (1708880, Bio-Rad). The assay is performed in accordance with the Minimum Information Required for Publication of Quantitative Real-Time PCR Experiments guidelines [44]. GAPDH was used as a reference control to standardize Ct values for individual genes. The relative expression level of each gene was analyzed according to 2−ΔΔCT [ΔΔCt = ΔCt (test) − ΔCt (calibrator)] method [45].

Table 1 qPCR primers used for gene expression analysisStatistical analyses

Statistical analyses and statistical graphs were conducted by GraphPad Prism 8 software. To compare the differences between the two variables, Student's t-test, Welch's correction, Mann–Whitney test were applied. Student's t-test was applied to data with Gaussian distribution and equal variance. Welch's correction was applied to data with Gaussian distribution and heteroscedasticity. Mann–Whitney test was applied to non-normally distributed data. Mouse survival curves were calculated using the Kaplan–Meier method and compared by the log-rank test. The data were presented as mean ± standard deviation (SD). All tests in this study were two sided. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 indicated the significant difference.